Skip To Main Content
 
  Figure 1.1 Incident & prevalent patient counts (USRDS), by modality
  Figure 1.2 Adjusted incident rates of ESRD & annual percent change
  Figure 1.3 Geographic variations in adjusted incident rates of ESRD per million population, 2010, by HSA
  Figure 1.4 Incident counts & adjusted rates of ESRD, by age
  Figure 1.5 Incident counts & adjusted rates of ESRD, by race
  Figure 1.6 Incident counts & adjusted rates of ESRD, by Hispanic ethnicity
  Figure 1.7 Incident counts & adjusted rates of ESRD, by primary diagnosis
  Figure 1.8 Adjusted incident rates of ESRD due to diabetes, by age, race, & ethnicity
  Figure 1.9 Adjusted incident rates of ESRD due to hypertension, by age, race, & ethnicity
  Figure 1.10 Adjusted prevalent rates of ESRD & annual percent change
  Figure 1.11 Geographic variations in adjusted prevalent rates of ESRD per million population, 2010, by HSA
  Figure 1.12 Prevalent counts & adjusted rates of ESRD, by age
  Figure 1.13 Prevalent counts & adjusted rates of ESRD, by race
  Figure 1.14 Prevalent counts & adjusted rates of ESRD, by Hispanic ethnicity
  Figure 1.15 Prevalent counts & adjusted rates of ESRD, by primary diagnosis
  Figure 1.16 Incident patient distribution, by first modality & payor
  Figure 1.17 Prevalent patient distribution, by modality & payor
  Figure 1.18 Vascular access use at initiation & on day of eligibility, 2010
  Figure 1.19 Access use at first outpatient hemodialysis, by pre-ESRD nephrology care, 2010
  Figure 1.20 Mean hemoglobin at initiation, by pre-ESRD ESA treatment
  Figure 1.21 Variations in the % of patients initiating dialysis with hemoglobin <10 g/dl, 2010
  Figure 1.22 Patient distribution at initiation, by eGFR (ml/min/1.73 m2)
  Table 1.a Patient demographics & adjusted rates, by ESRD network: incident dialysis patients, 2010
  Table 1.b Patient demographics & adjusted rates, by ESRD network: December 31 point prevalent dialysis patients, 2010
  Table 1.c Patient demographics & adjusted rates, by ESRD network: December 31 point prevalent transplant patients, 2010
  Table 1.d Incident counts & adjusted rates of ESRD at initiation, day 90, &one year, by modality, age, gender, race, ethnicity, & primary diagnosis, 2010
  Table 1.e Prevalent counts & adjusted rates of ESRD, by modality, age, gender, race, ethnicity, & primary diagnosis, 2010
  Table 1.f Pre-ESRD nephrologist care (column percent), 2010
  Table 1.g Percent of patients initiating dialysis with laboratory values outside the test’s normal limit, by age, gender, race, ethnicity, & primary diagnosis, 2010
View
Download
Chapter (zip)*
Slides (zip)*
*corresponding data in Excel included
Search This Page
Search All
Translate

Chapter 1

Incidence, prevalence, patient characteristics, & modality

Introduction

In 2010, the number of new patients starting therapy on hemodialysis declined for the first time in more than three decades. The population initiating on peritoneal dialysis, in contrast, grew for the second year in a row, and now accounts for 6.6 percent of patients with a known dialysis modality. This change may foreshadow those to come under the new bundled payment system, with its clear incentives for this form of home dialysis. Total incident dialysis cases rose 0.27 percent in 2010, to 114,083, while 2,863 patients received a preemptive transplant as their first ESRD modality; a total of 116,946 patients began ESRD therapy in 2010.

The rate of new ESRD cases per million population, which has been relatively stable since 2000, fell 2.0 percent in 2010, to 348. Growth continues to be driven by a relatively linear increase in the number of patients age 45–64; growth in the population age 65 and older, in contrast, has slowed considerably, but a slight upturn is present among those age 65–74, which could reflect the emergence of the post-World War II baby boomers into retirement age.

The incidence of ESRD in the black/African American population has finally started to decline, overall and for ESRD due to diabetes. Among those age 20–39, however, differences between whites and blacks/African Americans continue to be dramatic, with rates among the latter up to 3.8 times greater. Rates are also considerably higher for blacks/African Americans age 60 and older than for their white counterparts, though the gap is beginning to narrow.

The December 31, 2010 prevalent population included 383992 patients on hemodialysis and 29,733 on peritoneal dialysis, as well as 179,361 with a functioning kidney transplant; the total treated ESRD population thus rose to 593,086 — growth of 4 percent from 2009, which is the smallest increase in 30 years. The rate of prevalent ESRD cases reached 1,752 per million population, an increase of 1.1 percent from 2009, and also the slowest growth in the last three decades.

Insurance coverage in the dialysis population continues to change, with more incident dialysis patients now covered by Medicare Advantage. Private insurance, in contrast, is dominant among patients who receive a preemptive kidney transplant. In the 2010 prevalent population, 84 percent of hemodialysis patients and 79 percent of those on peritoneal dialysis had some type of Medicare coverage, compared to just 65 percent of those with a transplant.

Nephrology care prior to ESRD continues to be a concern. Since the 2005 introduction of the new Medical Evidence form (2728), with fields addressing pre-ESRD care, there has been little progress made in this area (pre-ESRD data, however, should be interpreted with caution because of the potential for misreporting). Forty-three percent of new ESRD patients in 2010, for example, had not seen a nephrologist prior to beginning therapy. And among these patients, 88 percent of those on hemodialysis began therapy with a catheter, compared to 54 percent of those who had received a year or more of nephrology care. Among those with a year or more of pre-ESRD nephrologist care, in contrast, 26 percent began therapy with a fistula — eight times higher than the rate among non-referred patients.

Data on patient care at the start of ESRD therapy show that the percentage of patients receiving an erythropoiesis stimulating agent (ESA) prior to initiation continues to decline, reaching just 20 percent in 2010 compared to one-third in the early part of the decade. This may reflect concern over potential adverse events when hemoglobin levels are targeted to a level above 12 g/dl. The mean hemoglobin at initiation of ESRD treatment is now 9.73 g/dl. These changes place different demands on care after the initiation of dialysis, and may alter the likelihood of a patient receiving a blood transfusion. The balance between cardiovascular risk with a hemoglobin greater than 12 g/dl and the risk of transfusion with lower hemoglobin levels needs to be addressed by patients and their physicians, particularly in the case of patients contemplating a kidney transplant, for whom sensitization from blood transfusions is to be avoided if at all possible.

The percentage of dialysis patients beginning therapy with an estimated glomerular filtration rate (eGFR, calculated with the CKD-EPI formula) above 15 ml/min/1.73 m2 continues to rise. It is not clear if this progressive increase is the result of severe comorbidity or a simple numerical starting point based on the ability to calculate the eGFR. Hopefully, symptoms and complications of uremia are still the primary indications for starting renal replacement therapy rather than a simple number, one which has been brought into question in recent years in controlled trials of early versus later dialysis initiation.

Biochemical data, collected on the Medical Evidence form since 2005, show that 57 percent of new patients in 2010 had an albumin less than the lower limit of normal, and 55 percent had a hemoglobin lower than 10 g/dl. Total cholesterol was greater than 200 mg/dl in 16 percent of patients, while 28 percent had an LDL level greater than 100 mg/dl, and 58 percent had an HDL level less than 40 mg/dl. Among patients with diabetes, 28 percent had a hemoglobin A1c level greater than 7 percent.

Recent changes and new incentives in the bundled prospective payment system for dialysis patients, introduced in January, 2011, may alter several characteristics of the incident and prevalent populations — particularly, due to cost incentives, the mix of peritoneal dialysis and hemodialysis patients. It is unclear how the expansion of peritoneal dialysis will affect patient outcomes, and how the new incentives will impact the emerging daily home hemodialysis population; provider incentives for this therapy are less clear, particularly as related to training. A more detailed assessment of the bundled payments is presented in Chapter Ten, and in future ADRs the USRDS will continue to assess the impact of this payment system on the ESRD population. Incident & December 31 point prevalent ESRD patients.

Figure 1.1 Incident & prevalent patient counts (USRDS), by modality

Incident counts &rates

Figure 1.2 Adjusted incident rates of ESRD &annual percent change

After a 1.1 percent increase in 2009, the adjusted incident rate of end-stage renal disease fell 2.0 percent in 2010, to 348 per million population. Since 2000, changes in rates have shown little variation, ranging from -2.1 percent to 2.1 percent. Incident ESRD patients. Adjusted: age/gender/race; ref: 2005 ESRD patients.

Figure 1.3 Geographic variations in adjusted incident rates of ESRD per million population, 2010, by HSA

In 2010, the adjusted incident rate of ESRD was 348 per million population, averaging 471 in the upper quintile. The highest adjusted rates occur in the Ohio Valley, portions of Texas and California, and the southwestern states. (Rates are not adjusted for ethnicity.) Incident ESRD patients. Adjusted: age/gender/race; reference: 2005 ESRD patients.

Table 1.a Patient demographics & adjusted rates, by ESRD network: incident dialysis patients, 2010

With an overall rate for incident dialysis patients of 340 per million population in 2010, rates by network range from 230 in Network 16 to 411 in Network 8. The distribution of patients by race continues to vary widely across the country. Blacks/African Americans, for example, constitute just 6.3 percent of the new dialysis population in Network 16, but 50–55 percent of patients in Networks 6 and 8. Incident dialysis patients. Adjusted: age/gender/race; reference: 2005 patients. Zero values in this cell.

Figure 1.4 Incident counts & adjusted rates of ESRD, by age

Since 2000, the adjusted incident rate of ESRD has grown 12.2 percent for patients age 75 and older, to 1,773 per million population in 2010, while rates for those age 0–19 and 20–44 have increased 9.1 and 6.3 percent, respectively, to 15.5 and 128. Rates for patients age 45–64 and 65–74, in contrast, though rising slightly during the decade, are now 5.3 and 3.1 percent lower than in 2000, at 581 and 1,368 per million, respectively.Incident ESRD patients. Adjusted: gender/race (1.4); reference: 2005 ESRD patients.

Figure 1.5 Incident counts & adjusted rates of ESRD, by race

By race, rates for blacks/African Americans and Native Americans in 2010 were 924 and 465 per million population, respectively — 3.4 and 0.5 times greater than the rate of 276 found among whites. Since 2000, the rate of new ESRD cases has grown 6.1 percent among whites and 2.5 percent among Asians, while decreasing 7.0 percent in the black/African American population. Adjusted: age/gender (1.5); reference: 2005 ESRD patients.

Figure 1.6 Incident counts & adjusted rates of ESRD, by Hispanic ethnicity

Thirteen percent of new ESRD patients in 2010 were Hispanic, a rate unchanged from those of the previous two years. While their rate of ESRD fell 1.7 percent, to 501, it remains 1.5 times greater than that seen among non-Hispanics. Adjusted: age/gender (1.6); reference: 2005 ESRD patients.

Figure 1.7 Incident counts & adjusted rates of ESRD, by primary diagnosis

With the exception of an uptick in 2006, the rate of new ESRD cases caused by diabetes has remained quite stable since 2000, reaching 152 per million population in 2010. The rate of ESRD due to hypertension, while down 2.2 percent in 2010, is 7.7 percent higher than the 2000 rate, at 99, while the rate of ESRD due to glomerulonephritis has fallen 27 percent, to 22.7. Incident ESRD patients. Adj:usted age/gender/race (1.7); reference: 2005 ESRD patients.s

Incident rates &racial differences

Figure 1.8 Adjusted incident rates of ESRD due to diabetes, by age, race, & ethnicity

Both the rates of incident ESRD caused by diabetes and their growth over time continue to vary widely by age and race/ethnicity. Among whites age 30–39, for example, the incident rate (adjusted for gender) has fallen just 1.0 percent since 2000, and in 2010 was 35.4 per million population. For blacks/African Americans of the same age, in contrast, the rate has increased 69 percent since 2000, to reach 133.8. The Native American population has seen a rise of 30.1 percent for this age group over the same time period, reaching 116 per million in 2010. And while rates of new ESRD cases among Asians remain comparatively low, among those age 30–39 they have nearly doubled since 2000, reaching 32.6 per million population in 2010.

Different patterns are seen among older populations. Among whites age 60–69, the rate of incident ESRD due to diabetes has fallen 3.6 percent since 2000, in contrast to a 29 percent increase in those age 70 and older. In blacks/African Americans, the rate for those age 60–69 has fallen 17.2 percent since 2000, while rates have decreased 40.4 and 18.4 percent, respectively, in Native Americans age 60–69 and those 70 and older. The rate for Hispanics age 60–69 has fallen 15.7 percent since 2000, to 1,166 in 2010, but has now surpassed the 2010 rate of 1,138 found in Native Americans of the same age. Incident ESRD patients; rates are three-year rolling averages. Adjusted: gender; reference: 2005 ESRD patients.

Figure 1.9 Adjusted incident rates of ESRD due to hypertension, by age, race, & ethnicity

As with diabetic ESRD, there are significant disparities by age, race, and ethnicity in the incidence of ESRD due to hypertension. Among whites age 30–39, for example, the rate per million population (adjusted for gender) rose 78 percent to between 2000 and 2010, to reach 15.3. The rate for blacks/African Americans of the same age rose at a far slower pace of 11.8 percent, but reached nearly 164 per million population — nearly 11 times greater than that of their white counterparts. Between 2000 and 2010, rates rose 19.8 and 9.2 percent in whites and blacks/African Americans age 70 and older, to reach 554 and 1,597. The rate increased 7.6 percent among Native Americans of the same age, reaching 420, but fell almost 22 percent in those age 60–69, to 143.6. The rate for blacks/African Americans age 60–69 was 955 per million population in 2010, 6.2 percent higher than the rate of 155 found in whites of the same age. Incident ESRD patients; rates are three-year rolling averages. Adjusted: gender; reference: 2005 ESRD patients.

Prevalent counts &rates

Figure 1.10 Adjusted prevalent rates of ESRD & annual percent change

The adjusted rate of prevalent cases of end-stage renal disease rose 1.7 percent in 2010 — slightly lower than the 1.9 percent growth in 2009 — to 1,763 per million population. This rate is 21 percent higher than that seen in 2000. The annual rate of increase has remained between 1.7 and 2.3 percent since 2004. » Figure 1.10; see page 429 for analytical methods. December 31 point prevalent ESRD patients. Adjusted: age/gender/race; reference: 2005 ESRD patients.

Figure 1.11 Geographic variations in adjusted prevalent rates of ESRD per million population, 2010, by HSA

In 2010, the rate of prevalent ESRD was 1,752 per million population. Patterns generally follow those found in the incident population, with an additional pocket of higher rates in the Dakotas and Minnesota. Rates in the upper quintile average 2,457. (Rates are not adjusted for ethnicity.) » Figure 1.11; see page 429 for analytical methods. December 31 point prevalent. patients Adjusted: age/gender/race; reference: 2005 ESRD patients.

Table 1.b Patient demographics & adjusted rates, by ESRD network: December 31 point prevalent dialysis patients, 2010

In 2010, the overall rate for December 31 point prevalent dialysis patients was 1,218 per million population. The percentage of prevalent dialysis patients with ESRD caused by diabetes ranges from 40 in Networks 1, 5, and 10 to 52–53 in Networks 14 and 15. December 31 point prevalent dialysis patients. Adjusted: age/gender/race; reference: 2005 patients. “.” Zero values in this cell.

Table 1.c Patient demographics & adjusted rates, by ESRD network: December 31 point prevalent transplant patients, 2010

For December 31, 2010 point prevalent transplant patients, the adjusted rate per million population is lowest in Network 6, at 428, and greatest in Network 11, at 789. As in the incident population, racial disparities persist. In Network 6, for example, blacks/African Americans account for 67 percent of prevalent dialysis patients, but only 39.6 percent of the prevalent transplant population. December 31 point prevalent transplant patients. Adjusted: age/gender/race; reference: 2005 patients. “.” Zero values in this cell.

Figure 1.12 Prevalent counts & adjusted rates of ESRD, by age

Reaching 6,068 per million population in 2010, the adjusted rate of prevalent ESRD for patients age 65–74 has increased 27 percent since 2000, while the rate among those age 75 and older has grown 44 percent, to 5,865. Among those age 20–44 and 45–64, in contrast, growth has been 14 and 19 percent, respectively, to 940 and 3,402 per million. December 31 point prevalent ESRD patients. Adjusted: gender/race (1.12); reference: 2005 ESRD patients.

Figure 1.13 Prevalent counts & adjusted rates of ESRD, by races

By race, rates of prevalent ESRD remain greatest in the black/African American and Native American populations, at 5,242 and 2,566 per million population in 2010, compared to 1,311 and 2,101 among whites and Asians.December 31 point prevalent ESRD patients. Adjusted:age/gender (1.13): ref: 2005 ESRD patients.

Figure 1.14 Prevalent counts &a djusted rates of ESRD, by Hispanic ethnicity

Rates of ESRD among Hispanics reached 2,606 in 2010, 1.5 times greater than that in the non-Hispanic population.

December 31 point prevalent ESRD patients. Adjusted; age/gender (14); reference: 2005 ESRD patients.

Figure 1.15 Prevalent counts & adjusted rates of ESRD, by primary diagnosis

Rates of ESRD due to diabetes and hypertension rose 1.8 and 2.1 percent, respectively, in 2010, to 656 and 437 per million population. ESRD caused by cystic kidney disease rose 1.8 percent, to 85 per million, and ESRD due to glomerulonephritis remained stable, at 263. December 31 point prevalent ESRD patients. Adjusted: age/gender/race (1.15); ref: 2005 ESRD patients.

Incident & prevalent modality

Table 1.d Incident counts & adjusted rates of ESRD at initiation, day 90, &one year, by modality, age, gender, race, ethnicity, &primary diagnosis, 2010

In 2010, 103,874 new patients began ESRD therapy on hemodialysis, 7,586 were placed on peritoneal dialysis, and 2,572 received a preemptive transplant (these data exclude patients with missing demographic information). Rates per million population were 316, 23.3, and 7.9, respectively. Past studies have suggested high mortality and significant movement between modalities in the first 90 days after ESRD initiation. The total number of 2010 incident patients with a known modality fell 11.4 percent between initiation and day 90. The hemodialysis population at day 90 was 13 percent smaller than at initiation; the peritoneal dialysis and transplant populations, in contrast, gained 5 and 22 percent, respectively. Between initiation and day 90, the rate per million population for hemodialysis fell from 316 to 273, while the rate for transplant rose from 7.9 to 9.7, and that for peritoneal dialysis rose from 23.3 to 24.6.At one year, the total number of patients with a known modality was 16.6 percent smaller than at day 90, and 26 percent smaller than at initiation. The rate per million population fell to 220 for hemodialysis, 20.8 for peritoneal dialysis, and 15.7 for transplant. » Table 1.d; see page 429 for analytical methods. Incident ESRD patients, 2010; unknowns dropped. Adjusted: age, gender, race. Reference: 2005 ESRD patients.

Figure 1.16 Incident patient distribution, by first modality & payor

Forty-five percent of new hemodialysis patients in 2010 were covered solely by Medicare, 14 percent had dual Medicare/Medicaid coverage, and 15.6 percent were covered by a Medicare HMO provider. Medicare covered 42 and 22 percent of new peritoneal dialysis and transplant patients, while 10.2 and 4.3 percent were dually-enrolled, and 10.3 and 3.9 percent had HMO coverage. Incident ESRD patients; peritoneal dialysis consists of CAPD & CCPD only.

Table 1.e Prevalent counts & adjusted rates of ESRD, by modality, age, gender, race, ethnicity, & primary diagnosis, 2010

On December 31, 2010, more than 376,000 ESRD patients were receiving hemodialysis therapy, 29,267 were being treated with peritoneal dialysis, and 174,136 had a functioning graft. Rates of ESRD in the prevalent population continue to be highest among blacks/African Americans, at 4,109 per million population for hemodialysis, 199.5 for peritoneal dialysis, and 925.6 for transplant. Prevalent rates for Asian patients on peritoneal dialysis or with a transplant are higher than those of their Native American counterparts. At 1,896, however, the rate of Native Americans receiving hemodialysis is 46 percent greater than that found in the Asian population, and more than double that found in whites. December 31 point prevalent ESRD patients, 2010; unknowns dropped. Adjusted: age, gender, race. Reference: 2005 ESRD patients

Figure 1.17 Prevalent patient distribution, by modality & payor

Nine of ten prevalent hemodialysis patients had some type of Medicare coverage in 2010, with 39 percent covered solely by Medicare, and 32 percent under Medicare/Medicaid. In the transplant population, in contrast, nearly one-third were covered solely by Medicare. Transplant patients younger than 65 and not disabled lose their entitlement after three years with a functioning graft. Coverage by non-Medicare insurers continues to increase in the dialysis population, in 2010 reaching 10.7 and 10.0 percent for hemodialysis and peritoneal dialysis patients, respectively. December 31 prevalent ESRD patients; peritoneal dialysis consists of CAPD & CCPD only.

Patient Characteristics

Table 1.f Pre-ESRD nephrologist care (column percent), 2010

Forty-three percent of patients starting ESRD therapy in 2010 had not seen a nephrologist prior to initiation. Of these patients, 89 percent initiated with a catheter and only 3 percent with a mature fistula; 13 percent had a maturing internal access. Patients with more than one year of pre-ESRD nephrologist care, in contrast, were far more likely to initiate with a mature fistula, at 26.3 percent. Incident ESRD patients, 2010.

Figure 1.18 Vascular access use at initiation & on day of eligibility, 2010

Data from the Medical Evidence form indicate that nearly 80 percent of 2010 incident hemodialysis patients initiated treatment with a catheter as their vascular access, 16.3 percent started with an arteriovenous (AV) fistula, and 3.2 percent initiated with an AV graft. By month four (day 91) of treatment, claims data show rates of catheter, AV fistula, and AV graft use were 52.6, 16.7, and 5.9 percent, respectively. Incident hemodialysis patients, July–December, 2010.

Figure 1.19 Access use at first outpatient hemodialysis, by pre-ESRD nephrology care, 2010

Among hemodialysis patients who have seen a nephrologist for more than a year prior to starting ESRD therapy, 41.8 percent initiate treatment using a catheter; these patients have the greatest likelihood at initiation of having an arteriovenous fistula (AV) or maturing fistula, at 31.3 and 20.1 percent, respectively. Patients with no pre-ESRD nephrology care most frequently start treatment with a catheter, at 81 percent, while only 18.4 percent initiate with either a mature or maturing AV fistula or graph. Incident hemodialysis patients, 2010.

Figure 1.20 Mean hemoglobin at initiation, by pre-ESRD ESA treatment

In the incident ESRD population, the mean hemoglobin at initiation has continued to fall from its peak in 2006, reaching 9.73 g/dl overall, 9.76 for patients receiving pre-ESRD treatment with an erythropoiesis stimulating agent (ESA), and 9.71 for patients without ESA treatment. At the end of 2010, 20 percent of new patients had received a pre-ESRD ESA. Incident ESRD patients.

Figure 1.21 Variations in the % of patients initiating dialysis with hemoglobin <10 g/dl, 2010

The percentage of patients initiating dialysis with a hemoglobin less than 10 g/dl is highest in parts of Texas and states along the Gulf Coast and Atlantic Seaboard, averaging 58.4 percent in the upper quintile. Incident ESRD patients.

Table 1.g Percent of patients initiating dialysis with laboratory values outside the test’s normal limit, by age, gender, race, ethnicity, & primary diagnosis, 2010

eGFR: ml/min/1.73 m2; serum albumin < lab lower limit.

*A1c data include only patients with diabetes as their primary diagnosis or as a comorbidity.

The likelihood of starting dialysis with laboratory values outside the normal limit is, with few exceptions, similar across demographic and disease categories. Overall, 56.9 percent of patients start treatment with a serum albumin below the test’s lower limit, and 55 percent have a hemoglobin less than 10 g/dl. Sixteen percent initiate with a total cholesterol greater than 200 mg/dl, 28.3 percent have low density lipid (LDL) measurements more than 100 mg/dl, and 58 percent have high density lipid (HDL) levels below the Adult Treatment Panel (ATP) III target of 40 mg/dl. Triglyceride levels above 150 mg/dl occur in 38.3 percent of incident patients, and 28 percent have a glycosylated hemoglobin (A1c) level above the recommended maximum of 7 percent. Incident ESRD patients, 2010.

Figure 1.22 Patient distribution at initiation, by eGFR (ml/min/1.73 m2)

Comparisons of estimated glomerular filtration rates (eGFRs) at the initiation of ESRD therapy indicate that patients are starting treatment sooner than in the past. In 2010, 29 percent initiated treatment with an eGFR of 10–<15 ml/min/1.73 m2, compared to 17.7 percent in 2000. And 16 percent started with an eGFR of 15 or greater, in contrast to 7.4 percent in 2000. Incident ESRD patients.